How to Do Research with The Archive • Zettelkasten Method
How to Do Research with The Archive • Zettelkasten Method
My very simple setup: Create entry points to help you do your actual work.
Howdy, Stranger!
How to Do Research with The Archive • Zettelkasten Method
Comments
I agree.
I also realized that when I use Zettelkasten and study a topic, it is easier to collect content in a structure note and then extract the note into an atomic note when the content is complete rather than immediately following atomicity. This will prevent you from writing duplicate content, and it will also make it easier to build the structure from scratch.
I think of this as the crystallization of metal. https://pocketdentistry.com/1-4-structure-of-metals-and-alloys/ When metal is melted, it is liquid. As it solidifies, crystal nuclei are gradually formed. When grain boundaries are formed, they are separated from other areas. In Zettelkasten, the moment grain boundaries are formed is the moment when the content is separated into atomicity and revealed. At that point, extracting an atomic note is good from a cognitive point of view.
If you try to follow atomicity too extremely from the beginning,
1) it can be difficult to see the whole picture,
2) because you are looking at things too decomposed, you can lose the characteristic of looking at the whole in a Gestalt way.
Interesting topic/challenge. We may have a different definition of research.
For the part of research that is “thinking”, “concept development”, “classification”, and “logic/outline” I fully agree. The Archive (or any Zettelkasten) works great here. In My experience it mainly requires discipline to make the structurenotes, the links, etc. My largest personal challenge here is to cooperate with others, those unfortunates who do not have not want to have any Zettelkasten, and still send you Word documents, etc.
However there are other aspects to research, for example “making a literature overview”, “collecting data” (either qualitative or quantitative), and “analyzing data”. I have tried to use my Zettelkasten for this but have not yet found a naturel or efficient workflow. Spreadsheets still rule here.
@erikh said:
Here I think what can be helpful is first thinking more abstractly in terms of data models and methods, instead of first reaching for a particular type of software.
Doing a literature review, for example, is collecting and analyzing a kind of qualitative data, so it is not too surprising that, as I mentioned before (quoting Joel Chan), there are "researchers who repurpose qualitative data analysis [QDA] tools like NVivo and Atlas.ti to do literature reviews".
And as I mentioned elsewhere, Ryan Murphy described how he used Obsidian for a particular qualitative analysis project. Chris Grieser read Murphy's article and was inspired to make an Obsidian plugin called Quadro to better handle a common QDA coding workflow. What Grieser did is a good example of someone who started by thinking in terms of a common data model and method, and then built the software that he wanted, in his case a plugin for the personal knowledge base software app Obsidian, which of course is also commonly used for keeping a Zettelkasten. Now that The Archive supports plugins, perhaps he could have done the same for The Archive.
All of this is a way of agreeing that basic Markdown-based personal knowledge base (Zettelkasten) software apps are not going to provide the best workflow for all problems "out of the box", but through plugins and interfacing with other software, a more integrated software system may be possible than you have yet discovered!
Agreed with @Andy Every aspect of research has its own best apps and there is no good in trying to force-fit anything.
I think we should be careful if we say “doing research with The Archive” that we qualify in which aspects of research it shines (and then share practices) and where it is not appropriate.
Again, in my experience, for conceptualization of research (half the work!), The Archive (or any Zettelkasten - I am paper-based now) works quite well.
With qualitative analysis there are possibilities, and it is these possibilities that I am most interested in. I have been trying for example to do qualitative literature analysis with the Zettelkasten and found it to work to a certain extent, but I found it difficult to find a good workflow. Same thing for a case study (which can look like a “world” as has been discussed in threads on writing fiction with the Zettelkasten): while I see how this could work, I could not find a good practical way. Both these qualitative analysis examples by the way do not have other existing “best practices workflows” nor any dedicated software solutions.
For any quantitative or even semi-quantitative analysis I have not found the Zettelkasten to work, and of course there are many dedicated research procedures and software packages. I do not really see the need to try and do anything with The Archive or any other Zettelkasten here. I checked out the example of Ryan Murphy, but have to say it looks clumsy to me. To my knowledge tools like Nvivo and Atlas read source documents also in pdf or MS-word format (which is normally the format you get them in), so there’s no need to convert everything to .md format. A bit like we used to print it all out 25 years ago. Also the functionality with the codes is already part of such software. Of course you could emulate that functionality with obsidian or even with links or tags, but it seems clumsy.
@erikh: I mostly agree with what you said except for the analogy that converting to Markdown is "like we used to print it all out 25 years ago". This is not at all true for well-structured formats, including MS-word format, which are very easily converted to Markdown using Pandoc.
On the other hand, a PDF that is not semantically tagged is a barbaric format that is indeed about as good as having a printout from 25 years ago. I don't have experience converting such PDFs to a structured format, but after all the advances in AI, I would not be surprised if that conversion is easy now too using some current tool that I don't know about.
I've never used NVivo or Atlas.ti, but they are much more expensive than Obsidian with the Quadro plugin. For some users, the latter could be sufficient and more cost-effective, but I don't know. What's interesting to me is that people are at least trying it.